Hi

I am fitting a model like this:

Code:
webuse nlswork, clear

mixed ln_w tenure || id: tenure, cov(unstruct) // Intercept-slope covariance, or

mixed, stddev // Intercept-slope correlation
Note the intercept-slope correlation of -.0185. How do I get that parameter from the model?

I realized that the covariance I can get with:

Code:
local covaria = round(tanh(_b[atr1_1_1_2:_cons]) * ///
                       exp(_b[lns1_1_1:_cons]) * ///
        exp(_b[lns1_1_2:_cons]), .000001)
di `covaria'
But how can I get the correlation? It's probably totally trivial but I can't get my head around it.

So many thanks
Go